We have undertaken a project to generate models of human congenital defects by screening ENU-mutagenized mice for recessive mutations affecting late embryonic development. The screen incorporated a genetic mapping component, with an aim to facilitate the positional cloning and functional characterization of the mutant genes. The strategy has worked well, and we have generated many mutant lines with phenotypes similar to human malformation syndromes and birth defects. The spectrum of abnormalities found to date is remarkably varied; for example, we have generated models of spondylocostal dysostosis, Robin sequence, congenital diaphragmatic defect, non-syndromic cleft palate, polycystic kidney disease, epidermal bullosa, non-bullosa congenital icthyosiform erythroderma, and structural heart disease. We have mapped a number of these, and identified the mutated locus in 9 lines. To accomplish this, we have taken advantage of efficent technologies for genetic mapping and positional cloning. The functions for many of the genes we have identified are not well understood, and we have initated a variety of biochemical and developmental studies to explore them. In addition to characterizing the biology of the defects in the mutant mice, we have in several cases established that the genes we identified play a role in the causation of human disease. Thus, all of the premises that were the basis of our original proposal have been experimentally validated. In this continuation proposal we hope to refine the specificity and sensitivity of the screen and optimize several aspects of the analysis, while maintaining the fundamental approach that has thus far proven so productive.